Dysregulated sarcoplasmic reticulum calcium release: Potential pharmacological target in cardiac disease

Abstract

In the heart, Ca²⁺ released from the intracellular Ca²⁺ storage site, the sarcoplasmic reticulum (SR), is the principal determinant of cardiac contractility. SR Ca²⁺ release is controlled by dedicated molecular machinery, composed of the cardiac ryanodine receptor (RyR2) and a number of accessory proteins, including FKBP12.6, calsequestrin (CASQ2), triadin (TRD) and junctin (JN). Acquired and genetic defects in the components of the release channel complex result in a spectrum of abnormal Ca²⁺ release phenotypes ranging from arrhythmogenic spontaneous Ca²⁺ releases and Ca²⁺ alternans to the uniformly diminished systolic Ca²⁺ release characteristic of heart failure. In this article, we will present an overview of the structure and molecular components of the SR and Ca²⁺ release machinery and its modulation by different intracellular factors, such as Ca²⁺ levels inside the SR as well as phosphorylation and redox modification of RyR2s.We will also discuss the relationships between abnormal SR Ca²⁺ release and various cardiac disease phenotypes, including, arrhythmias and heart failure, and consider SR Ca²⁺ release as a potential therapeutic target.

1. Introduction

Despite significant advances in treatments, cardiac diseases remain the leading cause of mortality and mor!bidity in Western nations. Development of new efficient treatment strategies requires understanding of basic mechanisms underlying cardiac function. Since the classical experiments of Dr. Sydney Ringer, Ca²⁺ has been recognized as a principal agent governing the mechanical activity of the heart (Ringer, 1883). On a beat-to-beat basis, Ca²⁺ derived from the intracellular Ca²⁺ storage site, the sarcoplasmic reticulum (SR), results in activation of contractile filaments and shortening of cardiac myocytes. Alterations in timing and magnitude of SR Ca²⁺ release result in different cardiac diseases such as arrhythmia and heart failure. SR Ca²⁺ release is controlled by specialized molecular machinery, composed of the cardiac ryanodine receptor (RyR2) and a number of accessory proteins, including FKBP12.6, calsequestrin (CASQ2), triadin (TRD) and junctin (JN). This multimolecular Ca²⁺ signaling complex is essential to normal Ca²⁺ cycling and is a site of potential pathologic failure. In this review, we will present an overview of the molecular organization of the Ca²⁺ release machinery and its modulation by different intracellular factors, such as Ca²⁺ levels inside the SR as well as phosphorylation and redox modification of RyR2s.We will also consider the relationships between abnormal SR Ca²⁺ release and various cardiac disease phenotypes, including, catecholaminergic polymorphic ventricular tachycardia, triggered and reentrant arrhythmias and heart failure. Finally we will consider SR Ca²⁺ release as a potential therapeutic target.

2. The sarcoplasmic reticulum

2.1. Structure and continuity

Striated muscle sarcoplasmic reticulum is a highly specialized form of endoplasmic reticulum (ER), which was described first in detail for both skeletal and cardiac muscle as an extensive network of interconnected cisternae and tubules throughout the sarcoplasm (Fawcett & McNutt, 1969; Sommer, 1995; Franzini-Armstrong & Protasi, 1997). Estimates of the total SR volume in ventricular myocytes of different mammalian species vary between 0.87 and 7.9% of cell volume, and it is commonly assumed to constitute 3.5% of cell volume (Bers, 2001). The SR consists of two distinct regions: junctional SR (jSR), which directly faces the transverse tubules, and free SR (fSR) which faces the myofibrils and does not form junctions with other membrane systems. See Fig. 1A and B. The total SR volume is divided approximately equally between jSR and fSR. In rabbit ventricular myocytes, jSR forms extended, flattened cisternae with an average diameter of 592 nm (Brochet et al., 2005). Each cistern carries 66 RyR2s clustered in one or two release units on the junctional membrane. The jSR contains an electron dense material which is formed by the Ca²⁺ binding protein calsequestrin (CASQ2) (Tijskens et al., 2003). The free SR is devoid of internal content, and its external surface exhibits densely distributed 8 nm particles representing the Ca²⁺ ATPase molecules. Although topologically continuous with the rest of the SR network, the connectivity of an average jSR cistern to fSR is limited to a few narrow passages; therefore, diffusional barriers for the free translocation of Ca²⁺ and other molecules exist between these compartments (Brochet et al., 2005). The restricted connections between jSR and fSR facilitates the generation of local luminal Ca²⁺ gradients, which are thought to play an important role in luminal Ca²⁺ signaling (Gyorke and Terentyev, 2008).

Fig. 1.

SR microanatomy and Ca²⁺ signaling. A. Diagram illustrating the relationship between the free SR (fSR) network, the junctional (jSR) cisternae, T tubules, feet and calsequestrin along with its anchoring filaments in a mammalian ventricular myocyte. B. Electron micrograph of junctional SR of a rabbit ventricular myocyte. Between the arrows RyR2s are seen as foot structures in the gap between jSR and T tubule membranes. C. Action potential-elicited cytosolic and intra-SR Ca²⁺ signals followed by caffeine-induced Ca²⁺ release and SR depletion. Upper: profiles and linescan images of cytosolic Ca²⁺ changes measured with Rhod-2; lower: profiles and linescan images of [Ca²⁺]_SR changes monitored with Fluo 5N. D. Simultaneous measurements of local SR Ca²⁺ release (Rhod-2, upper) and depletion signals (Fluo 5N, lower) (i.e. sparks and blinks). Panel B from DXP Brochet, D Yong, De Maio A, WJ Lederer, C Franzini-Armstrong, H Cheng. Ca²⁺ blinks: rapid nanoscopic store calcium signaling. PNAS 2005; 102: 3099–3104. Copyright (2005) National Academy of Sciences, U.S.A.

2.2. SR as a Ca²⁺ storage and release organelle

The Ca²⁺ stored in and released from the SR is a critical determinant of cardiac contractility. Additionally intra-SR Ca²⁺ levels play a key role in SR-mediated Ca²⁺ signaling by controlling the functional activity of RyR2s. In mammalian cardiac myocytes, the total SR Ca²⁺ content has been estimated to be in the range of 50–150 µmol/ l in the cytosol (or 1.4–3 mmol/l of SR volume), considering that SR comprises 3.5% of cell volume (Bers, 2001). Due to the presence of substantial amounts of Ca²⁺ binding sites in the SR, only a fraction of this Ca²⁺ is free. According to various estimates, 50 to 90% of the total Ca²⁺ is bound to Ca²⁺ buffering proteins such as CASQ2 (Shannon & Bers,1997; Shannon et al., 2000), and the existence of large reserves of bound Ca²⁺ permits the SR to serve as an efficient intracellular Ca²⁺ storage site despite its minute space. Whereas bound Ca²⁺ contributes to the storage capacity of the SR, free Ca²⁺ plays an important role in store operation by setting thermodynamic limits on Ca²⁺ATPase Ca²⁺ transport activity, establishing the driving force for SR Ca²⁺ release, and modulating RyR2 activity. As measured directly by low affinity Ca²⁺ dyes loaded into the SR, the resting (diastolic) free [Ca²⁺]SR constitutes 1 to 1.5 mM and becomes only partially depleted (24% to 63%) during contraction (Shannon et al., 2003a). See Fig. 1C and D. Local depletion signals in individual SR cisternae during Ca²⁺ sparks (termed “blinks”) have been measured in myocytes from both normal and diseased hearts (Brochet et al., 2005; Kubalova et al., 2005). The occurrence of local SR Ca²⁺ depletion is consistent with restricted diffusion of Ca²⁺ from fSR to jSR (described above). Indeed, a rapid refilling of local jSR Ca²⁺ store from the bulk of the SR would be expected to prevent significant local depletion. Interestingly, as determined by luminal Ca²⁺ measurements with Fluo5N, Ca²⁺ diffusion through the SR network along the length of the myocyte is unexpectedly fast (60 µm² s⁻¹,(Guo et al., 2007). This reinforces the notion that diffusion barriers between jSR and the rest of the SR network must exist to explain jSR Ca²⁺ depletion during localized Ca²⁺ release events.

3. The Ca²⁺ release channel complex

The cardiac Ca²⁺ release channel is a macromolecular complex composed of more than a dozen of proteins centered around the RyR2 (Franzini-Armstrong & Protasi, 1997; Zhang et al., 1997; Marx et al., 2000; Bers, 2004). See Fig. 2. The cardiac RyR2 is composed of four homologous subunits of 560KD that are encoded by a separate gene of three existing genes (Nakai et al., 1990; Fill & Copello, 2002). While a significant portion of the RyR2 extends into the SR lumen, the bulk of the RyR2 homotetramer is situated on the cytosolic side of the SR membrane (visible on EM micrographs and described as the “foot”); it carries multiple sites for cytosolic regulatory molecules (Ca²⁺, ATP, etc.) as well as docking sites for accessory regulatory proteins (FKBP12.6, calmodulin, sorcin, etc.). On the luminal side, the RyR2 is associated with a number of proteins, including TRD, JN and CASQ2. TRD and JN bind directly to RyR2 and each other and mediate interactions of CASQ with RyR2. The structure of RyR2 at an atomic resolution remains to be resolved. Structure–function studies, and analysis of mutational hot spots revealed several key regions, referred to as domains (Ikemoto & Yamamoto, 2002; George et al., 2007). Interactions between these domains (domain “unzipping”) are thought to underlie transitions between conformations corresponding to open and closed states of the channel, as well as transduction of regulatory events, and pathologic influences from both the cytosolic and luminal side.

Fig. 2.

Schematic presentation of the RyR2 channel complex. The domain structure of RyR2 with sites of interaction with auxiliary proteins and phosphorylation sites. Two major divergent regions located in the cytosolic portion are indicated. Calsequestrin, junctin, and triadin, proteins interacting with RyR and between themselves and RyR2 in the SR, are shown. PP, protein phosphatase; P, phosphorylation sites; CaM, calmodulin; CaMKII, calmodulin-dependent protein kinase II. Adapted from Gyorke and Terentyev, 2008.

3.1. FK binding protein 12.6

FK binding proteins (FKBPs) are immunophilins, and are so named based on binding to the immunosuppressant tacrolimus (originally named FK506) (Brillantes et al., 1994). In cardiac muscle, FKBP12 and FKBP12.6 are present (Jayaraman et al., 1992; Timerman et al., 1993; Timerman et al., 1996), and FKBP12.6 has been put forward as a key modulator of RyR2 function in both health and disease (Marx et al., 2000; Wehrens et al., 2005a; Yano et al., 2005). However, the role of FKBP12.6 in regulating RyR2 function remains highly controversial. FKBP12.6 is preferentially associated with RyR2 due to higher affinity (Timerman et al., 1996). The binding stoichiometry is one FKBP12.6 per RyR subunit, but the FKBP12.6 binding site in the RyR2 linear sequence has not been precisely identified. Dissociation of FKBP12.6 has been reported to result in increased RyR2 activity. It has been suggested that PKA phosphorylation at 2809 results in dissociation of FKBP12.6 from the RyR2 with a subsequent increase in the open probability of the RyR2 (Brillantes et al., 1994; Ahern et al., 1994; Kaftan et al., 1996). However, other studies have failed to detect any changes in RyR2 function upon dissociation of FKBP12.6 (Barg et al., 1997; Xiao et al., 2007). Furthermore, it has been reported that FKBP12.6 remains attached to RyR2 following phosphorylation (Stange et al., 2003; Xiao et al., 2004).

3.2. Calsequestrin

Calsequestrin (CASQ) is an intra-SR protein which has a low affinity and high-capacity for Ca²⁺ binding (MacLennan & Wong, 1971; Campbell et al., 1983; Mitchell et al., 1988; Beard et al., 2004; Gyorke & Terentyev, 2008). In the SR, calsequestrin exists as a mixture of monomers, dimers, and various multimers. When Ca²⁺concentrations exceed 1–10 mmol/l, CASQ monomers and dimers form linear oligomers, and these polymers (Park et al., 2004) can be detected as electron-dense filamentous matrices both in vitro and in vivo (Somlyo et al., 1981; Saito et al., 1984; Franzini-Armstrong et al., 1987). CASQ polymers are primarily located in the junctional SR (jSR) in both skeletal and cardiac muscle (isoforms CASQ1 and CASQ2, respectively) (Jorgensen & Campbell, 1984). Due to the proximity of CASQ matrices to RyR2, CASQ2 polymers are thought to provide the stores of releasable Ca²⁺ which activate myocyte contraction. In this regard, condensed CASQ represents the high Ca²⁺ binding capacity form of the protein and possesses at least twice the Ca²⁺ binding capacity of monomeric CASQ in both skeletal and cardiac muscle (Tanaka et al., 1986; Park et al., 2004). CASQ condensation reportedly involves head-to-tail oligomerization that occurs through front-to-front and back-to-back dimerization contacts involving the N-and C-terminal regions of the protein, respectively (Wang et al., 1998; Park et al., 2004). While the precise mechanism(s) of Ca²⁺ sequestration by CASQ is not known, it has been suggested that CASQ polymers provide electrostatically charged surfaces for absorption of Ca²⁺ (Wang et al., 1998).

In addition to serving as a Ca²⁺ storage site, cardiac calsequestrin (CASQ2) has also been shown to regulate RyR2 activity more directly via protein–protein interactions involving triadin (TRD) and junctin (Jn) (Gyorke et al., 2004; Gyorke & Terentyev, 2008; Terentyev et al., 2007). Both the N-terminal (Zhang et al., 1997; Kobayashi et al., 2000) and C-terminal regions of CASQ2 have been suggested as the triadin (and/or junctin) interaction sites (Shin et al., 2000). It has been proposed that CASQ2mediates the responsiveness of the RyR2 channel to luminal Ca²⁺ by serving as a luminal Ca²⁺ sensor (Gyorke et al., 2004; Terentyev et al., 2007; Gyorke & Terentyev, 2008). In particular, we demonstrated that both triadin 1 and junctin exert potentiatory influences on RyR2 activity (Gyorke et al., 2004). Thus, at low luminal Ca²⁺, CASQ2 is tightly bound to triadin 1 (junctin) and prevents it from activating the RyR2 channel. When luminal Ca²⁺ is increased, the binding of Ca²⁺ to CASQ2 disrupts the interactions of CASQ2with TRD thereby relieving the inhibitory influence of CASQ2 on the RyR2 channel. Thus, luminal Ca²⁺-dependent termination of Ca²⁺ release could be partly attributed to inhibition of the RyR2 by CASQ2 at lowered [Ca²⁺]_SR. At the same time, when there is an increased [Ca²⁺]_SR, the increase in the fraction of CASQ2-uninhibited RyR2 channels could contribute to the known enhancement of diastolic SR Ca²⁺ leak in cardiac myocytes. Modulation of channel function seems to be done by the monomer as inhibition occurs at low concentrations when CASQ2 exists as a monomer. In summary, CASQ2 appears to play at least two different roles in cardiac myocytes: as a Ca²⁺ storage reservoir in the SR, and as an active modulator of the Ca²⁺ release process, mediated by monomers.

3.3. Triadin and junctin

Triadin (TRD) and junctin (JN) are transmembrane proteins in the junctional SR that appear to bind directly to the RyR2, and to CASQ2 thereby anchoring CASQ2 to the RyR2 channel (Jones et al., 1995; Kobayashi & Jones, 1999; Kobayashi et al., 2000; Gyorke & Terentyev, 2008). Three isoforms of triadin (35, 40 and 92 kD) have been described of which triadin 1 (40 kD) is the predominant, representing more than 95% of the cardiac protein (Guo et al., 1996; Kobayashi & Jones, 1999). Junctin was identified as a 26-kDa CASQ2 binding protein in cardiac and skeletal muscle (Jones et al., 1995). Triadin 1 and junctin exhibit significant structural similarities. Both proteins are composed of a single membrane spanning domain, a short cytoplasmic N-terminal segment and a long, highly positively charged C-terminal domain situated in the lumen of the SR. The luminal domains of both proteins are characterized by the frequent occurrence of long stretches of alternating positively and negatively charged residues known as KEKE motifs (Jones et al., 1995; Kobayashi et al., 2000). These regions, in general denoted as protein–protein binding motifs, are thought to be involved in interactions between triadin 1 and junctin themselves and in interactions with RyR2 and CASQ2 (Zhang et al., 1997). In lipid bilayers both TRD and JN activate RyR2s [Gyorke 2004], and acute overexpression of TRD also results in increased RyR activity (Terentyev et al., 2005). The functions of TRD1 and JN have also been studied using genetic mouse models (Zhang et al., 2001; Kirchhefer et al., 2001; Kirchhefer et al., 2003; Kirchhefer et al., 2006; Yuan et al., 2007). However, the interpretation of the physiologic roles of TRD and JN from the transgenic models is complicated by various adaptive and pathological changes (e.g. hypertrophy and heart failure) accompanying chronic protein expression. Our current understanding, as described above (Section 3.2) is that the complex of CASQ, TRD and JN serves to regulate RyR2 activity in a [Ca²⁺]_SR-dependent manner.

3.4. Other proteins

A number of other proteins interact with the RyR2 complex, including sorcin (Valdivia, 1998), homer (Pouliquin et al., 2006; Worley et al., 2007), and the histidine-rich Ca²⁺ binding protein (Kim et al., 2003). The specific role of these proteins in SR Ca²⁺ release regulation remains to be defined.

4. Modulation of sarcoplasmic reticulum Ca²⁺ release

4.1. Modulation of sarcoplasmic reticulum Ca²⁺ release by luminal Ca²⁺

Rather than merely serving as a passive pool of available Ca²⁺, intra-SR Ca²⁺ plays an active role in controlling the Ca²⁺ release process. With measurements of cell-wide Ca²⁺ release, this regulatory influence manifests itself as an increase in the fraction of total releasable Ca²⁺ from the SR at elevated Ca²⁺ loads and conversely as a decrease of fractional release at reduced SR Ca²⁺ loads (Bassani et al., 1995; Shannon et al., 2000). Increasing SR Ca²⁺ content also enhances the frequency of spontaneous Ca²⁺ release events, Ca²⁺ sparks (Cheng et al., 1996; Lukyanenko et al., 1996). The effects of luminal Ca²⁺ on SR Ca²⁺ release have been attributed to twomechanisms: 1) Activation of RyR2s on the cytosolic side of the channel at sites involved in CICR by Ca²⁺ passing through the channel; and 2) stimulation of RyR2s activity at distinct sites on the luminal side of the channel. The relative contribution of each mechanism to Ca²⁺ signaling in myocytes is difficult to define due to significant interdependence of the two mechanisms.

The stimulatory effects of luminal Ca²⁺ on RyR2 channels have been directly demonstrated in in vitro reconstitution studies(Lukyanenko et al., 1996; Xu & Meissner, 1998; Gyorke & Gyorke, 1998; Ching et al., 2000; Gyorke & Terentyev, 2008). Consistent with the existence of Ca²⁺ regulatory sites on both sides of the channel, two different mechanisms of action of luminal Ca²⁺ upon RyR2 have been described: 1) indirect, “feed-through” activation by luminal Ca²⁺ at the cytosolic activation site of the channel; and 2) direct, “true luminal” activation by Ca²⁺ on the luminal side of the channel. In single reconstituted RyR2s, these two mechanisms are readily distinguishable based on differences in the voltage-dependency of luminal Ca²⁺ effects and the dissimilar impact of luminal Ca²⁺ on channel gating behavior which are characteristic of these two mechanisms. For example, feed-through effects are characteristically larger at negative holding potentials than at positive holding potentials due to the reliance of the stimulatory action of luminal Ca²⁺ on luminal-to-cytosolic fluxes which are intrinsic to this mechanism (Xu & Meissner, 1998). Furthermore, feed-through activation increases RyR2 open probability by prolonging the lifetime of channel openings rather than by increasing their frequency. This is understandable considering that luminal Ca²⁺ can flow through the RyR2 channel and reach the cytosolic activation sites only when the channel is open; once the channel closes, Ca²⁺ accumulated near the cytosolic face of the RyR rapidly dissipates rendering the channel inactive. Compelling evidence for true luminal Ca²⁺ activation of RyR2s has been obtained under experimental conditions that prevent any luminal-to-cytosolic Ca²⁺ fluxes (measurements performed at highly positive holding potentials and millimolar cis [Ca²⁺]) and by using tryptic digestion of the luminal phase of the RyR2 channel to disable the putative luminal Ca²⁺ sensor (Gyorke & Gyorke,1998; Ching et al., 2000).

4.2. Modulation of the Ca²⁺ release channel by protein kinases and phosphatases

In addition to being governed by changes in cytosolic and SR luminal Ca²⁺, the RyR2 channels are targets for modulation by a number of intracellular ligands and signaling cascades, including phosphorylation pathways. The RyR2 is phosphorylated by PKA and CAMKII at Ser-2809 and by CAMKII at Ser-2815 (Marx et al., 2000; Wehrens et al., 2004). Recently a new PKA phosphorylation site has been reported at Ser-2830 (Xiao et al., 2005; Xiao et al., 2006), although this site has been disputed (Wehrens et al., 2006). In addition to these identified phosphorylation sites, RyR2 is phosphorylated by CAMKII on as many as five sites, based on ³²P phosphate incorporation measurements (Rodriguez et al., 2003). The RyR2 is reportedly preassociated with PKA, and the protein phosphatases PP1 and PP2A through the anchoring protein mAKAP (Marx et al., 2000). In addition, RyR2 co-immunoprecipitates with CAMKII (Zhang et al., 2003) and PP2B (calcineurin, CN, (Bandyopadhyay et al., 2000), However, the molecular sites for these interactions remain to be identified. Thus, the molecular components involved in both phosphorylation and dephosphorylation of the RyR2 are anchored directly to the protein.

Despite clear evidence for RyR2 phosphorylation, the physiological consequences of this process remain highly controversial. PKA-dependent phosphorylation has been shown to either increase or decrease steady-state RyR2 open probability in different reports (Bers & Guo, 2005; Benkusky et al., 2004; Wehrens et al., 2004). Marks and collaborators have reported that PKA phosphorylation at Ser-2809 leads to dissociation of FKBP12.6 from the RyR2, in turn resulting in enhanced channel activity (Marx et al., 2000); however, other investigators find that FKBP12.6 still binds to RyR2 after PKA phosphorylation (Stange et al., 2003; Xiao et al., 2004). In intact myocytes, PKA stimulation has been shown to enhance SR Ca²⁺ release, an effect that could be attributed to phosphorylation of PLB and increased SR Ca²⁺ load as well as increased I_Ca. Interestingly, in PLB knock-out myocytes, RyR2 phosphorylation by PKA produced no effects on spontaneous local Ca²⁺ release signals [i.e. Ca²⁺ sparks, (Li et al., 2002)], indicating that the PKA-dependent increase in Ca²⁺ spark frequency is entirely attributable to PLB phosphorylation and the resulting increase in SR Ca²⁺load. A recent study showed that RyR2 is stimulated only during extreme hyperphosphorylation (i.e. all 4 subunits of the tetramer are phosphorylated at Ser-2809) (Carter et al., 2006). CAMKII, expressed as the delta isoform, has also been shown to produce disparate effects. Some authors have reported that exogenous CaMKII increases RyR2 activity (Witcher et al., 1991; Hain et al., 1994), whereas others have found that exogenous (Lokuta et al., 1995) and endogenous CaMKII reduces RyR2 opening or ryanodine-RyR binding (Takasago et al., 1991). Similarly, in intact voltage clamped myocytes some investigators have reported that both endogenous and exogenous CAMKII increase the amplitude of Ca²⁺ transients at a given SR Ca²⁺ content and I_Ca trigger (Bers & Guo, 2005), whereas others have reported that endogenous CAMKII and constitutively active CAMKII suppress Ca²⁺ transients and Ca²⁺ sparks (Wu et al., 2002; Yang et al., 2007). Thus, this remains an open area of investigation at the present time.

The action of protein kinases is opposed by dephosphorylating phosphatases. Three types of protein phosphatases, including PP1, PP2A and PP2B (calcineurin), have been shown to influence cardiac performance (Neumann, 2002). Overall, in the majority of studies, phosphatases reportedly downregulate SR Ca²⁺ release and contractile performance (Neumann et al., 1993; duBell et al., 1996, 2002; Carr et al., 2002; Santana et al., 2002). Application of PP1 and PP2A to permeabilized myocytes caused an acute activation of Ca²⁺ sparks followed by depletion of the SR Ca²⁺ content (Terentyev et al., 2003a), suggesting that depression of EC coupling by phosphatases may involve depletion of the SR Ca²⁺ stores due to increased activation of RyR2s. Consistent with this possibility, phosphatases have been shown to increase RyR2 activity in some (Lokuta et al., 1995; Terentyev et al., 2003a; Carter et al., 2006) but not all single channel studies (Hain et al., 1994). Overall, as summarized in this brief overview, the mechanisms and functional consequences of RyR2 phosphorylation remain to be conclusively elucidated.

4.3. Redox modulation

Ryanodine receptor channels are subject to modifications by oxidizing and reducing agents capable of interacting with thiolcontaining cysteine residues of the channel protein. The cardiac RyR contains approximately 90 cysteines per subunit, of which approximately 20 are in reduced state (Xu et al., 1998). These free cysteines are targets for several types of redox modifications, including disulfide crosslinking, S-nitrosylation and S-glutathionylation. In general, oxidizing conditions increase RyR2 activity and therefore stimulate SR Ca²⁺ release (Abramson & Salama, 1988; Eager et al., 1997; Zima & Blatter, 2006), while these effects are reversed by reducing reagents such as dithiothreitol (DTT). Although different classes of oxidizing agents tend to increase RyR2 activity, the underlying effects on channel gating are not uniform. For example, whereas oxidation increases the sensitivity of the channel to Ca²⁺ activation, S-glutathionylation decreases the sensitivity of the channel to inhibition by Mg²⁺ (Donoso et al., 2000). There is also evidence of time-dependent effects whereby prolonged exposure to high doses of oxidizing agents, after an initial stimulation of activity, can cause irreversible inactivation of RyRs (Eager et al., 1997; Marengo et al., 1998). Thus, the functional consequences of RyR2 thiol modification are dependent on the concentration of the oxidizing agent, exposure time and the nature of the chemical reaction involved in the modification. These different effects are likely to reflect participation of different sets of cysteine residues under different oxidizing conditions. The role of specific cysteine residues in RyR redox modifications has only started to be elucidated. It has been shown that redox modification involves only 12 of the total 100 cysteines on RyR1 (Aracena-Parks et al., 2006). Whereas some of these cysteines are non-specific targets for different redox agents, other sites are exclusively S-nitrosylated (Cys-1040, Cys-1303) or S-glutathionylated (Cys-1591 and Cys-3193). Endogenous oxidizing agents (reactive oxygen species or ROS) are continuously generated in the heart and are likely to exert tonic influences on intracellular targets/processes including SR Ca²⁺ release. ROS production has been shown to increase in various pathological settings such as ischemia-reperfusion and heart failure. Growing evidence suggests that modification of RyR2 by ROS contributes to the abnormal Ca²⁺ handling associatedwith these disease states (see section 6.7 below).

5. The Ca²⁺ release-uptake cycle

5.1. Activation of sarcoplasmic reticulum Ca²⁺ release

Myocardial excitation–contraction (EC) coupling begins with membrane depolarization activating voltage-dependent Ca²⁺ channels in the plasma membrane which allows a relatively small amount of Ca²⁺ to enter the cell. This Ca²⁺ serves as a trigger signal to activate Ca²⁺ release channels present in large clusters in the junctional sarcoplasmic reticulum (jSR). See Fig. 3A. The release of Ca²⁺ via groups of RyR2s has been visualized as Ca²⁺ sparks (Cheng et al., 1993). Sparks have been estimated to involve coordinated openings of 6 to 30 RyR2s (Bridge et al., 1999; Lukyanenko et al., 2000;Wang et al., 2001). Ca²⁺ sparks amplify the initial Ca²⁺ influx trigger signal and combine to produce an elevation of cell-wide myoplasmic [Ca²⁺], the Ca²⁺ transient. This increase in cytosolic Ca²⁺ concentration ([Ca²⁺]_i) leads to activation of the contractile proteins and hence to generation of the heartbeat. Under certain conditions, Ca²⁺ release can be triggered by other mechanisms such as Ca²⁺ entry via reverse mode NCX (Sipido et al., 1997). Moreover, Ca²⁺ entry via these Ca²⁺ transport mechanisms can interact synergistically resulting in activation of large SR Ca²⁺ release at positive membrane potentials where I_Ca by itself is small. (Viatchenko-Karpinski et al., 2005; Sobie et al., 2008).

Fig. 3.

Ca²⁺ cycling in normalmyocytes andmyocytes under three different disease conditions: SR Ca overload, CPVT and heart failure. A. In normalmyocytes, Ca²⁺ influx through L-type Ca²⁺ channels (LCC) during the action potential activates RyR2s and initiates the release of Ca²⁺ stored in the SR on a beat-to-beat basis. Diastolic SR Ca²⁺ release is prevented due to RyR2 deactivation at reduced [Ca]_SR; [Ca]_SR is always below the threshold for activation of spontaneous Ca release (indicated by dashed line). B. Under conditions of Ca²⁺ overload, [Ca]_SR reaches the threshold for activation of spontaneous Ca release, diastolic SR Ca²⁺ release occurs, resulting in arrhythmogenic DADs through stimulation of NCX. C. Abnormal RyR2 channel function caused by CPVT-linked mutations in RyR2 and CASQ2 result in a reduced threshold (indicated by red dashed line) for spontaneous Ca²⁺ release (i.e. perceived Ca²⁺ overload), diastolic Ca²⁺ release and DADs. D. In heart failure, acquired defects in RyR2 result in a more severe RyR2 deregulation ([Ca²⁺]_SR threshold for spontaneous Ca²⁺ release is further reduced); excessive SR Ca²⁺ leak leads to depletion of the [Ca]_SR and diminished systolic Ca²⁺ release and contraction.

5.2. Termination of sarcoplasmic reticulum Ca²⁺ release

Following its rapid activation, SR Ca²⁺ release robustly terminates leaving a substantial Ca²⁺ reserve of releasable Ca²⁺ in SR. The exact underlying causes of Ca²⁺ release termination remain to be defined (Stern & Cheng, 2004). Several mechanisms have been proposed, including inactivation of the RyR2 channel by elevated cytosolic Ca²⁺ (i.e. Ca²⁺-dependent inactivation); obligatory shift from high- to low modes of activity (i.e. adaptation); and RyR2 deactivation caused by decline in intra-SR [Ca²⁺] (i.e., luminal-Ca²⁺-dependent deactivation). In recent years, several lines of evidence have pointed to a prevailing role of luminal Ca²⁺ in controlling SR Ca²⁺ release (Terentyev et al., 2002; Sobie et al., 2002; Zima et al., 2008; Terentyev et al., 2008). For example, buffering [Ca²⁺]SR using low affinity exogenous Ca²⁺ buffers such as citrate or maleate prolonged the duration of Ca²⁺ release during both Ca²⁺ sparks and global Ca²⁺ transients (Terentyev et al., 2002). Subsequently, similar results were obtained by varying the expression level (increasing or decreasing) of the endogenous Ca²⁺ buffer CASQ2 in cardiac myocytes (Terentyev et al., 2003b). Recently, the role of this mechanism has received further support from the observation that Ca²⁺ sparks terminate at a constant free [Ca²⁺]_SR regardless of the extent of [Ca²⁺]_SR buffering by CASQ2 (Terentyev et al., 2008). Moreover, mutations in CASQ2 that enhance the responsiveness of RyR2s to luminal Ca²⁺ reduced the [Ca²⁺]_SR threshold for local Ca²⁺ release termination. Thus, mounting experimental evidence suggests that luminal Ca²⁺-dependent changes in RyR2s contribute to SR Ca²⁺ release termination in cardiac myocytes.

5.3. Refractoriness of Ca²⁺ signaling

After activation, a shift in the gating mode of RyR2s closes the channels, which then remain closed in a time-dependent manner. There is substantial evidence indicating that changes in luminal Ca²⁺ govern not only termination of CICR, but also induce a refractory state that persists until recovery of [Ca²⁺]_SR. (Terentyev et al., 2002, 2003b; Szentesi et al., 2004; Sobie et al., 2005). Thus, experimental interventions to slow SR-Ca²⁺ reuptake (either by SERCA-pump inhibitors or by increasing the intra-SR Ca²⁺ binding sites with either exogenous or endogenous buffers) slow recovery of Ca²⁺ signaling from refractoriness (Szentesi et al., 2004; Terentyev et al., 2002, 2003b). Moreover, accelerating [Ca²⁺]_SR recovery following Ca²⁺ release either by stimulation of the SERCA pump or by reducing endogenous CASQ levels accelerated recovery of RyR2 refractoriness. Interestingly, restitution of refractoriness appears to lag behind [Ca²⁺]_SR recovery (Wang et al., 2001; Sobie et al., 2005) suggesting that the relationship between recovery and luminal Ca²⁺ is non-linear and/or may involve some intermediate steps. Consistent with the latter possibility, modulation by luminal Ca²⁺ may involve a group of intermediate proteins including CASQ2, triadin and junctin (see Section 3.2 above).

5.4. Sarcoplasmic reticulum Ca²⁺ reuptake

During relaxation, rapid reuptake of Ca²⁺ into the SR is achieved by the SR Ca²⁺ ATPase (SERCA2a). The activity of SERCA is modulated by the phosphoprotein phospholamban (PLB) in a phosphorylation-dependent manner (Brittsan & Kranias, 2000). Dephosphorylated PLB inhibits SERCA activity whereas phosphorylation relieves the inhibition. Phosphorylation occurs at two sites, Ser 16 and Thr 17 by PKA and CAMK2 respectively. Within the SR a large fraction of Ca²⁺ is bound to CASQ, and this binding accelerates uptake by keeping free [Ca²⁺]_SR at low levels. The ability of SERCA to attain a maximum thermodynamically allowable [Ca²⁺]_SR is limited by the leak though RyRs which increases as a function of [Ca²⁺]_SR as described below (Lukyanenko et al., 2001).

5.5. Luminal Ca²⁺-dependent leak

The SR exhibits a substantial diastolic Ca²⁺ leak in the form of spontaneous Ca²⁺ sparks. The frequency of sparks and the amount of leak is increased at elevated SR Ca²⁺ content and conversely is reduced at lowered SR Ca²⁺ content (Lukyanenko et al., 2001; Shannon et al., 2002). This leak limits the ability of the SR Ca²⁺ pump to attain the maximum thermodynamically allowable [Ca²⁺]_SR and plays an important role in setting the SR Ca²⁺content. Mechanistically, the SR Ca²⁺ leak is a consequence of activation of RyRs by increased [Ca²⁺]_SR. Increased SR Ca²⁺ leak plays an important role in various pathological states such as HF, in which it results in reduced SR Ca²⁺ content and weakened contractility (see below).

5.6. Balance of trans-sarcolemmal and sarcoplasmic reticulum Ca²⁺ fluxes

In order to maintain Ca²⁺ homeostasis, an equivalent amount of Ca²⁺ must enter the myocyte via I_Ca and then be removed from the myocyte on a beat-to-beat basis. Most of this Ca²⁺ is removed by the electrogenic Na⁺–Ca²⁺ exchanger that exchanges three Na⁺ ions for one Ca²⁺ ion. At the same time, the amount of Ca²⁺ released from the SR must be balanced by the amount of Ca²⁺ taken back up by SERCA. Deviation from the balance of trans-sarcolemmal and SR Ca²⁺ fluxes will lead to either gain or loss of Ca²⁺ in the cell and in the SR. This requirement for the homeostatic balance of cellular and SRCa²⁺ fluxes is an important factor in determining the SR Ca²⁺ content and release, and in shaping the response of SR Ca²⁺ release to pharmacological and/or pathophysiologic modulation.

By studying the effects of low concentrations of the RyR2 channel modulators (caffeine and tetracaine) on systolic Ca²⁺ release, Eisner and coworkers demonstrated that the SR Ca²⁺ release mechanism in cardiac myocytes exhibits an ability for self-regulation which allows myocytes to maintain a consistent amplitude of systolic Ca²⁺ transients, even when interventions alter RyR2 activity (Overend et al., 1998; Trafford et al., 2000). They attributed this self-regulation to a feedback loop consisting of SR Ca²⁺ release causing cellular Ca²⁺ extrusion; Ca²⁺ extrusion influencing SR Ca²⁺ load; and SR Ca²⁺ load affecting SR Ca²⁺ release (Eisner et al., 1998; Trafford et al., 2002). As susequently shown, the dependence of RyR gating on luminal Ca²⁺ is another important factor in determining Ca²⁺cycling, and is expected to both accelerate and increase the extent of restitution of Ca²⁺ transient amplitude after a change caused by a perturbed open probability of RyR2s (Lukyanenko et al., 2001; Shannon et al., 2002). It is to be noted, however, that the capacity of a myocyte to restore its initial Ca²⁺ transient amplitude following alterations in open probability of RyRs only operates over a limited range. Thus, if RyRs are hyperstimulated the SR becomes completely depleted and there is no SR Ca²⁺ release, and consequently only low amplitude steady-state Ca²⁺ transients. This situation occurs in certain pathological states such as HF and is examined in detail below.

6. Pathologies of cardiac sarcoplasmic reticulum Ca²⁺ signaling

6.1. Overview

Abnormal RyR2-mediated SR Ca²⁺ release has been implicated in a spectrum of cardiac disease states ranging from triggered and reentrant arrhythmias to HF (Wehrens et al., 2004; George et al., 2007; Ter Keurs & Boyden, 2007; Laurita & Rosenbaum, 2008). A classification of cardiac diseases related to abnormal SR Ca²⁺ release and their hypothetical relationship to defective function is presented in Fig. 4. Ca²⁺-dependent arrhythmias of both triggered and reentry types are generally associated with instability of SR Ca²⁺ release whereas HF is associated with diminished systolic SR Ca²⁺ release. In general, loss of control of CICR can arise in two ways. Firstly, in a variety of conditions that lead to increased SR Ca²⁺ load (i.e. “Ca²⁺ overload”), sensitization of RyR2s by enhanced luminal Ca²⁺ contributes to generation of spontaneous Ca²⁺ releases or instabilities of SR Ca²⁺ release that manifest themselves as Ca²⁺ alternans (Fig. 3B; Fig. 4). Whereas spontaneous Ca²⁺ releases causes DADs and triggered arrhythmia, Ca²⁺ alternans can provide basis for heterogeneities in electrical repolarization and substrate for reentry arrhythmia. Secondly, acquired or genetic defects in RyR2s enhance the responsiveness of RyR2 to luminal Ca²⁺ such that the channel becomes hyperactive even at reduced SR Ca²⁺ content. This state of “perceived” Ca²⁺ overload (Gyorke & Terentyev, 2008) can also lead to Ca²⁺-dependent arrhythmias and has been implicated familial tachyarrhythmias associated with mutations in RyR2 or CASQ2 (Fig. 3C). Diminished SR Ca²⁺ release (e.g. as in HF) can result either from failure of EC coupling or exhaustion of SR Ca²⁺ stores due to an excessive leak via RyR2s (Fig. 3D). Thus, different diseases could result from different degrees of alteration in the same underlying mechanism, namely increased RyR activity such that a relatively small change could result in destabilization of CICR whereas a more severe deviation could lead to SR Ca²⁺ leak and SR store depletion. Clearly, other factors besides abnormal RyR function, such as changes in SERCA and NCX activity, also contribute to different disease phenotypes. The role of these factors has been extensively reviewed (e.g. Kranias and Bers, 2007; Periasamy et al., 2008) and is not discussed here.

Fig. 4.

Classification of Ca²⁺-dependent disease states based on mechanisms of alterations of Ca²⁺ handling.

6.2. Ca²⁺ overload and triggered arrhythmia

Excess intracellular Ca²⁺ (i.e. Ca²⁺ overload) is a characteristic feature of various pathological states such as metabolic inhibition, ischemia/reperfusion and digitalis toxicity. Ca²⁺ overload is associated with triggered afterdepolarizations, which can initiate sustained tachyarrhythmias (Ferrier et al., 1973; Kass et al., 1978; Pogwizd & Bers, 2004; Maltsev et al., 2006; Venetucci et al., 2008). See Fig. 3B. The causal relationships between Ca²⁺ overload and triggered activity are known to include the following sequence of events: (1) SR Ca²⁺ overload leads to generation of spontaneous Ca²⁺ release from the SR, (2) elevated Ca²⁺ induces depolarizing membrane currents that give rise to delayed afterdepolarizations (DADs), (3) DADs activate ectopic action potentials, and (4) ectopic action potentials can then initiate tachyarrhythmias. A brief discussion of these steps follows below.

6.3. Spontaneous Ca²⁺ release

Since the initial studies of Fabiato in skinned myocytes (Fabiato & Fabiato, 1972), spontaneous Ca²⁺ release or Ca²⁺ waves have been described by many investigators in various preparations. Ca²⁺ waves in general are associated with increased SR Ca²⁺content (Stern et al., 1988; Pogwizd & Bers, 2004; Venetucci et al., 2008). Moreover, it has been shown with both total SR Ca²⁺ content determinations (Diaz et al., 1997) and more directly using SR-entrapped Ca²⁺ indicators (Kubalova et al., 2004) that Ca²⁺ waves arise when [Ca²⁺]_SR reaches a certain threshold value. See Fig. 3B. As visualized by confocal Ca²⁺ imaging, Ca²⁺ waves are initiated by local release events, Ca²⁺ sparks, that occur spontaneously in resting myocytes, and then propagate through saltatory activation of adjacent release sites (Cheng et al., 1996; Lukyanenko et al., 1999). The mechanisms for the transition from Ca²⁺ sparks to self-propagating Ca²⁺ waves have been examined. One possibility is that Ca²⁺ waves start and propagate when Ca²⁺ efflux via individual release sites becomes sufficient to raise Ca²⁺ at adjacent release sites above the threshold for CICR from a Ca²⁺-overloaded SR. However, this mechanism may not be sufficient by itself to account for Ca²⁺ wave generation. Experimental studies combined with mathematical modeling have suggested that propagation of Ca²⁺ waves relies on stimulatory effects of elevated luminal Ca²⁺ on RyR activity (Lukyanenko et al., 1999). As to the specific role of luminal Ca²⁺, two hypotheses have been considered: (1) increased luminal Ca²⁺ sensitizes the Ca²⁺ release channels to cytosolic Ca²⁺, thus enhancing the ability of cytosolic Ca²⁺ to activate adjacent release sites via CICR, and (2) Ca²⁺ transported from the wave-front into the adjacent SR elements raises luminal Ca²⁺ above a critical threshold level, resulting in activation of the release channels at a luminal site. Examination of the effects of SERCA inhibitors, such as thapsigargin, showed that Ca²⁺ waves elicited by local application of caffeine in myocytes preloaded with Ca²⁺ not only did not stop but even accelerated after rapid inhibition of SR Ca²⁺ uptake (Lukyanenko et al., 1999). At the same time, Ca²⁺waves could be elicited following sensitization of RyRs by agents such as caffeine — even in the absence of Ca²⁺ overload. These finding supported the role of CICR in wave propagation with luminal Ca²⁺ serving to sensitize RyRs. However, recently (Keller et al., 2007) reexamined the role of SERCA-mediated uptake in Ca²⁺ wave propagation using a photo-releasable SERCA inhibitor. Following inhibition of SR Ca²⁺ uptake these authors observed a reduction in Ca²⁺ wave velocity in contrast to previous findings. The reasons for these discrepant results remain undefined and additional experimentation will be required to resolve this issue. In summary, evidence exists to suggest that the effects of luminal Ca²⁺ on RyR2s play at least a sensitizing role in the generation of Ca²⁺ waves by promoting CICR. The intriguing possibility that [Ca²⁺]_SR plays a more direct triggering role in Ca²⁺ wave generation will require further verification.

6.4. Afterdepolarizations and triggered arrhythmia

Triggered arrhythmias originate from afterdepolarizations and are so named as they are “triggered” by either abrupt or sustained abnormal variations in heart rate. Afterdepolarizations occur either during the action potential (early afterdepolarizations or EADs) or after termination of the action potential (delayed afterdepolarizations or DADs). Classically, EADs are considered to occur at slow heart rates, DADs at rapid heart rates.

Early afterdepolarizations are often associated with delayed cardiac repolarization (action potential prolongation or QT/QTc prolongation on the electrocardiogram), and have been implicated in the genesis of Torsades de Pointes arrhythmias (see below). It has been suggested that repolarization abnormalities either directly cause or provide a foundation for EADs [reviewed by (Roden, 1993)]. Postulated ionic mechanisms responsible for EADs include K⁺ channel block, enhanced late sodium current, reactivation of L-type calcium current during prolonged action potential plateau. However, most relevant to this review, abnormal release of calcium from intracellular SR stores during repolarization can lead to inward current via Na⁺/Ca²⁺ exchange and activation of EADs, even during rapid heart rates. This has been demonstrated in computer simulations (Huffaker et al., 2004) and in isolated myocytes when SR Ca²⁺ is increased (Spencer & Sham, 2003). In addition to NCX, window I_Ca has also been implicated in the development of EADs; however, SR Ca²⁺ release has been implicated as an important modulator of I_Ca inactivation and EADs, in a canine model of TdP (chronic AV block) (Antoons et al., 2007).

Delayed afterdepolarizations are typically attributed to “calcium overload” of cardiac myocytes, with the classic examples being inhibition of Na⁺–K⁺-ATPase with cardiac glycosides (e.g. digoxin) or beta-adrenergic stimulation (Rosen & Legato, 1985; Rosen, 1985). DADs are attributed to spontaneous Ca²⁺ release from the sarcoplasmic reticulum (SR), resulting in subsequent activation of a transient inward current via the Na⁺/Ca²⁺ exchanger. Elevations of cytosolic Ca²⁺ resulting from spontaneous SR Ca²⁺ release causes inward current to flow across the sarcolemma, producing transient depolarizations or DADs (Lederer & Tsien, 1976; Allen et al., 1984; Wier & Hess, 1984; Marban et al., 1986). These inward currents are predominantly due to the electrogenic NCX exchanging 1 Ca²⁺ for 3 Na⁺ ion (Schlotthauer & Bers, 2000). In most cases DADs are not of sufficient amplitude to reach the threshold to initiate an AP (see Fig. 5). This is because a Ca²⁺ wave at any time occupies only a fraction of the cell; therefore only a fraction of the total NCX molecules are exposed to elevated Ca²⁺. However, when several waves arise in the cell simultaneously (i.e. multifocal waves), the summation of the inward currents can depolarize the cell to threshold resulting in an AP (Capogrossi et al., 1987). In intact muscle, currents can spread from cell to cell through the gap junctions, summating over a region covering a group of myocytes (area comparable in size to the electrical space constant). When spontaneous Ca²⁺ release occurs asynchronously among cells, the likelihood that this summation could reach threshold to trigger an ectopic AP is low. However, spontaneous Ca²⁺ releases can be relatively synchronized following an AP. Such temporal concurrence of spontaneous release events could be due to similarly timed recovery of RyR2s from luminal-dependent refractoriness and simultaneous attainment of the [Ca²⁺]_SR threshold for spontaneous Ca²⁺ release.

Fig. 5.

Substrate for re-entry and initiation of re-entry. A. Top panel: action potentials under control conditions. Middle panel: action potentials at a faster rate, with a delayed afterdepolarization (dashed line). Note that the delayed afterdepolarization does not reach the voltage threshold to trigger an action potential (threshold indicated by dotted line). Bottom panel: the delayed afterdepolarization is of sufficient amplitude to reach the threshold voltage and triggers an early action potential. B. The figure depicts an anatomic circuit which has the functional prerequisites to form a substrate for re-entry. Note that in pathway A, there is slowed conduction (represented by the slower upstroke of the action potential), while in pathway B, there is normal conduction but a prolongation of the action potential duration (and therefore prolongation of the refractory period) in the representative action potential. C. A properly timed premature beat excites the anatomic circuit and initiates a reentrant arrhythmia. Note that due to the slower conduction in pathway A, the impulse is very slowly conducted down pathway A. However, due to the longer refractory period which exists in pathway B, the premature beat is too early to excite pathway B, and conduction is blocked. As the premature impulse exits pathway A, the refractory period in pathway B has passed and the impulse excites pathway B in a retrograde manner. Due to the heterogeneities of conduction velocity and refractoriness, this anatomical circuit can maintain reentrant activation and thus, a tachyarrhythmia.

Afterdepolarizations of sufficient amplitude to reach the threshold voltage for a myocyte can initiate ectopic (triggered) beats. Arrhythmias can be induced by either isolated or sustained triggered beats. Notably, conduction of either EADs and DADs can give rise to ectopic (premature) beats which can initiate reentrant arrhythmias (see Fig. 5) resulting in sustained cardiac arrhythmias. Re-entry is thought to be the most common cause of sustained cardiac arrhythmias. Re-entrant arrhythmias require an underlying substrate; the substrate (predisposition) for re-entry is dispersion of both conduction velocities and refractory periods within a region of the heart. An illustration of the substrate and initiation of re-entry is shown in Fig. 5.

6.5. Catecholaminergic polymorphic ventricular tachycardia

Catecholaminergic polymorphic ventricular tachycardia (CPVT) is an inherited arrhythmia characterized by exercise or stress-induced syncope and sudden death (Eldar et al., 2002; Laitinen et al., 2004; Napolitano & Priori, 2007). CPVT accounts for approximately 20% of all sudden deaths that occur in the absence of structural heart disease. Two genetic forms of CPVT have been described: a prevailing dominant form linked to mutations in RyR2s and a rare recessive form associated with mutations in CASQ2 (Postma et al., 2002; Lahat et al., 2004). Currently about 60 mutations in RyR2 and 7 mutations in CASQ2 have been linked to CPVT. Mutations in these proteins account only for about 50% of all cases of CPVT suggesting that genetic defects in other proteins can also lead to this syndrome. Notably, these different defects result in the same arrhythmic phenotype; and experimental work has reconciled the mechanisms of CPVT from these varying mutations.

Adenoviral expression studies of CPVT-associated CASQ2 mutants in cardiac myocytes have demonstrated that various mutations act by inducing irregular Ca²⁺ transients and arrhythmogenic DADs (Terentyev et al., 2003b; Gyorke & Terentyev, 2008). The mechanisms by which CASQ2 mutations alter regulation of SR Ca²⁺ release have been examined (Terentyev et al., 2006; Gyorke & Terentyev, 2008). Although various CASQ2 mutations affect different aspects of CASQ2 function, the results converge in a single common pathway to shorten RyR2 refractoriness and promote spontaneous Ca²⁺ release in the diastolic phase. See Fig. 3C. First, as described above, CASQ2 polymers are required for high-capacity Ca²⁺ binding, and certain mutations (Del.339–354) appear to disrupt CASQ2 polymerization (Terentyev et al., 2008). Reduced Ca²⁺ binding capacity of the SR results in accelerated recovery of [Ca]_SR by the SERCA and thus premature functional recovery of RyR2 from luminal Ca²⁺-dependent deactivation. Secondly, interactions of CASQ2 with the RyR2 complex modulate the sensitivity of the RyR2 channel to luminal Ca²⁺. CASQ2 mutations [R33Q; (Terentyev et al., 2006)] alter the modulation of RyR2s by luminal Ca²⁺ without apparently altering the Ca²⁺ binding capacity of CASQ2. The resulting altered modulation of RyR2s by luminal Ca²⁺ (i.e. compromised ability of CASQ2 to deactivate the RyR2) with these mutations leads again to premature recovery of Ca²⁺ release from refractoriness.

The mechanisms by which CPVT-associated mutations in RyR2 alter RyR2 channel function have been investigated (George et al., 2007; Liu & Priori, 2008). Based on studies using a genetic mouse model, Marx and coworkers have suggested that RyR2s mutations lead to dissociation of the regulatory protein FKBP12.6 from RyR2 causing increased diastolic Ca²⁺ leak, DADs and arrhythmias (Wehrens et al., 2003). However, several aspects of this hypothesis have been disputed by other laboratories (George et al., 2007; Xiao et al., 2007; Liu & Priori, 2008). Alternatively, based on studies using a heterologous expression system, Chen and coworkers demonstrated that CPVT mutations increase the sensitivity of RyR2 to activation by luminal Ca²⁺ (Jiang et al., 2005). Similar to the effects of arrhythmogenic CASQ2 mutants, sensitization to luminal Ca²⁺ would reduce the [Ca²⁺]_SR threshold for the generation of spontaneous Ca²⁺ release events and DADs. Thus, abnormal luminal modulation may be a common mechanism for genetically-distinct forms of CPVT.

Based on accumulated data it appears that CPVT can arise as a result of a genetic defect in any component of the Ca-storage-release machinery involved in controlling, sensing and responding to changes in luminal Ca²⁺. The defects affect the ability of RyRs to stay refractory, thus resulting in abnormal diastolic Ca²⁺ release and DADs. In the context of this mechanism, the role of adrenergic stimulation in CPVT can be explained by PKA-dependent stimulation of SERCA, which would accelerate SR Ca²⁺ store reuptake. This would further enhance the rate and extent of recovery of release sites from luminal Ca²⁺-dependent refractoriness exacerbating the arrhythmogenic effects of CASQ2/RyR2 mutations. As to the sites of origin within the heart, a recent study using epicardial mapping and a genetic mouse model (RyR2/RyR2.R4496C) of CPVT reported that focal arrhythmias arise predominantly in the His-Purkinje system (Cerrone et al., 2007).

6.6. Ca²⁺ alternans

Altered intracellular Ca²⁺ homeostasis is not only involved in triggered activity but also may be part of pathogenesis of reentrant arrhythmias. Instabilities in Ca²⁺ handling have been directly linked to heterogeneities in conduction and repolarization providing a substrate for reentrant excitation. These instabilities often manifest themselves as beat-to-beat variations in contractility and AP duration referred to as cardiac alternans and its cellular equivalent Ca²⁺ alternans (Weiss et al., 2006; Eisner et al., 2006; Laurita&Rosenbaum, 2008).When alternans is initiated it usually occurs in an identical phase, i.e. long–short–long APs, in all regions of the tissue (concordant alternans).When the heart rate is faster than the critical rate, a shift of phase occurs such that some regions continue to alternate in the long–short–long pattern while other regions alternate in the reverse pattern (discordant alternans). Spatially discordant alternans are highly arrhythmogenic and greatly amplify the heterogeneity of repolarization and electrical refractoriness producing conduction block and a substrate for reentrant excitation (Pastore et al., 1999; Qu et al., 2000). Cardiac alternans occurs in various pathologic settings including ischemia, electrolyte disbalances and HF (Eisner et al., 2006; Laurita & Rosenbaum, 2008). Moreover, T-wave alternans (electrocardiographic manifestation of cardiac alternans), is recognized as a reliable predictor of VF and sudden cardiac death in patients with HF (Rosenbaum et al., 1994).

Initially, it was thought that cardiac alternans has its origin in a non-linear relationship between the diastolic interval and electrical repolarization, resulting in periodic changes in AP duration. However, more recent studies have suggested that Ca²⁺ alternans might be a cause rather than a consequence of AP alternans (Chudin et al.,1999; Huser et al., 2000; Diaz et al., 2004;Weiss et al., 2006; Eisner et al., 2006; Laurita & Rosenbaum, 2008). The influence of the Ca²⁺ transient on AP is attributable to effects on Ca²⁺ dependent currents, such as the NCX current and I_Ca, during the AP plateau. The coupling between Ca²⁺ and membrane potential (Vm) can be either positive or negative. When a large Ca²⁺ transient increases the duration of the AP via stimulating the inward NCX current (concordant electrical and mechanical activity), the Ca²⁺–Vm coupling is referred to as positive. When the large Ca²⁺ transient decreases the duration of AP (discordant electrical and mechanical activity), the Ca–Vm coupling is said to be negative.

In general, Ca²⁺ alternans is thought to arise when the rate of myocyte Ca²⁺ cycling cannot keep up with heart rate. In various pathological settings, slowed Ca²⁺ cycling pushes the heart into alternans even at normal heart rates. Thus, when SERCA Ca²⁺ transport activity is decreased the SR Ca²⁺ store can be refilled only on an alternating basis. Increasing SR Ca²⁺ buffering capacity by overexpressing CASQ2 or introducing low affinity exogenous buffers into the SR should slow functional recharging of the Ca²⁺ store and has also been shown to cause alternans. The steepness of the relationship between SR Ca²⁺ content and SR Ca²⁺ release is another important factor in the development of alternans (Weiss et al., 2006; Eisner et al., 2006). As discussed above, the steepness arises as a result of the combination of stimulatory effects of luminal Ca²⁺ on RyR function, and the presence of the positive feedback loop inherent to CICR. With a steep Ca²⁺ release–[Ca²⁺]_SR relationship, the SR will produce a large Ca²⁺ transient when the SR Ca²⁺ content is relatively high and a disproportionately small Ca²⁺ transient when the SR Ca²⁺ content is relatively low. An especially large Ca²⁺ transient will cause extrusion from the cell of an extra amount of Ca²⁺ resulting in an under-filled SR Ca²⁺ store and hence a smaller Ca²⁺ transient in the next cycle. This will decrease Ca²⁺ efflux and increase SR Ca²⁺ content to a larger extent, such that the Ca²⁺ transient will again be larger in the third cycle. In most disease conditions including acute ischemia and HF, alternans are thought to be caused by slowed Ca²⁺ cycling due to inhibition of SERCA (Weiss et al., 2006). However, we recently showed that in a post-infarction model of sudden cardiac death alternans is attributable to increased steepness of the Ca²⁺ release–[Ca²⁺]_SR relationship (Belevych et al., 2008). In this model abnormal SR Ca²⁺ release regulation is caused by redox modification of RyRs and can be attenuated by reducing agents.

Given the non-linear dependence of Ca²⁺ release on [Ca]_SR in the development of alternans, beat-to-beat variations in Ca²⁺ transients are expected to correlate with beat-to-beat variations in SR Ca²⁺ content. Indeed, such a correlation has been demonstrated experimentally by measuring both total SR Ca²⁺ content and free [Ca]_SR in myocytes exhibiting Ca²⁺ alternans (Diaz et al., 2004). However, other investigators failed to observe consistent changes in [Ca]_SR during alternans (Picht et al., 2006). These studies suggested that processes such as refractoriness control the number of available RyRs during each heart beat, and can produce alternans irrespective of changes in SR Ca²⁺ content. It is to be noted however, that RyR refractoriness itself appears to be controlled by [Ca]_SR (Terentyev et al., 2002; Sobie et al., 2005; Gyorke &; Terentyev, 2008). Thus alterations in the number of available RyRs still would be expected to correlate with alterations in diastolic Ca²⁺ transients. One potential explanation as to why this is not observed in some studies is that changes are too small to be detected. Alternatively, recovery occurs with a time lag following [Ca]_SR recovery, thus providing a window in which alternans can develop without significant alterations in diastolic [Ca]_SR.

6.7. Heart failure

Systolic heart failure (HF) is an increasingly prevalent disease and occurs when the cardiac muscle is too weak to pump sufficient blood through the body. Patients with HF typically die either due to ventricular arrhythmias or progressive failure of cardiac mechanical function (pump failure). A recent analysis of the mode of death in patients with heart failure (arrhythmias vs. pump failure) suggests that during earlier (less severe) stages of heart failure, death is seven times more likely to occur from cardiac arrhythmias, while in end-stage HF, death is twice is likely to occur from pump failure (Mozaffarian et al., 2007). Although the pathophysiologic mechanisms of HF are diverse, highly complex and not fully defined, reduced Ca²⁺ transient amplitude and RyR2 dysfunction with elevated SR Ca²⁺ leak are increasingly recognized features of the disease process (Marx et al., 2000; Shannon et al., 2003b; Kubalova et al., 2005). Uncontrolled RyR2 gating and increased diastolic SR Ca²⁺ leak have been shown to cause a reduction in the SR Ca²⁺ content, thus limiting the ability of cardiac muscle to contract in myocytes from failing hearts (Belevych et al., 2007). Additionally, abnormal RyR2 activity could contribute to the pathogenesis of arrhythmia in patients with HF.

Similar to CPVT, elevated RyR2-mediated SR Ca²⁺ leak in HF is associated with altered responsiveness of RyRs to luminal Ca²⁺ (Kubalova et al., 2005). See Fig. 3D. In particular, luminal Ca²⁺ dependence is markedly left-shifted in RyR2s from failing hearts. As discussed for CPVT, this specific defect may explain why Ca²⁺ leak remains high even at reduced SR Ca²⁺ load despite the well established positive relationship between RyR2 activity and SR Ca²⁺ content. Thus, acquired and genetic alterations in RyR seem to affect RyR behavior in a similar manner. While it is apparent that HF-associated RyR2-mediated SR Ca²⁺ leak could contribute to 1) depleted SR Ca²⁺ stores and reduced contractility and/or 2) Ca²⁺-dependent arrhythmias, the underlying mechanisms of RyR2 dysfunction remains a subject of intense debate.

One proposed mechanism for abnormal RyR behavior involves hyperphosphorylation of RyR2 by PKA (Marx et al., 2000; Wehrens et al., 2004). RyR2 phosphorylation at Ser-2809 reportedly causes dissociation of FKBP12.6 followed by increased RYR activity (see above). Another possibility is that increased RyR activity is caused by CAMKII phosphorylation at Ser-2815 (Ai et al., 2005). However, both the functional effects of phosphorylation on RyR function and the role of abnormal RyR phosphorylation in HF require further clarification. An alternative mechanism for enhanced RyR2 activity is modification of the channel protein by reactive oxygen and/or nitrogen species generated during HF. ROS levels are markedly elevated in failing hearts (Belch et al.,1991; Sawyer et al., 2002; Giordano, 2005) and redox modification is known to stimulate RyR activity (see above). Preliminary results obtained in our laboratory suggest that sulfhydryl oxidation of RyRs contributes to enhanced SR Ca²⁺ leak in chronic heart failure (Gyorke et al., 2007). Importantly, oxidation impairs the luminal Ca²⁺ dependence of RyRs in amanner characteristic of HF. At the same time, reducing agents normalize SR Ca²⁺ leak, RyR activity and luminal Ca²⁺ sensitivity. Additional studies are required to establish the relative contribution of this mechanism to HF-induced abnormalities. In summary, evidence obtained using variousmodels of HF and in human HF suggests that RyR2s become excessively active i.e., “leaky” in HF. This increase in RyR2 activity has been linked to defective modulation of the channel by SR luminal Ca²⁺, a mechanism that normally operates to terminate SR Ca²⁺ release and keep the RyR2 channels closed during diastole (see above). Modest alterations in this control mechanism destabilize CICR causing arrhythmogenic dispersion of Ca²⁺ signaling, whereas more severe defects result in a massive diastolic leak and consequently the depleted SR Ca²⁺ stores characteristic of heart failure.

6.8. Atrial fibrillation

Atrial fibrillation (AF) is the most common sustained cardiac arrhythmia. It has been estimated that the lifetime risk for the development of AF is 1 in 4 after reaching the age of 40 (Lloyd-Jones et al., 2004). Occasionally, AF occurs in the absence of detectable structural heart disease or any other recognizable risk factor. However, most commonly, AF occurs in the presence of known risk factors or diseases, including valvular heart disease, advanced age, hypertension, coronary artery disease, diabetes, hyperthyroidism, chronic obstructive pulmonary disease, and heart failure. AF is known to generally increase the risk of death, and specifically to increase the risk of death during heart failure.(Benjamin et al., 1998; Kammersgaard & Olsen, 2006) Recently, abnormal electrical activity originating in the sleeves of the pulmonary veins (PVs) has been implicated in the initiation of AF, particularly paroxysmal AF. However, non-PV atrial foci have also been clearly identified as a common cause of atrial arrhythmias in those with persistent AF(Haissaguerre et al., 2005a; Haissaguerre et al., 2005b). Thus, the initiation of AF is likely to be a multi-factorial process. Abnormalities in calcium handling have recently been implicated in the pathogenesis of AF. In atrial tissues from patients with chronic AF there are reported reductions in message for L-type Ca²⁺ channels and SERCA, most commonly without changes in message for phospholamban, CASQ, NCX or RyR2 (VanWagoner et al.,1999; Van Gelder et al.,1999; Brundel et al.,1999; Lai et al.,1999). However, RyR messagewas reportedly reduced in one study, and accompanied by reduced binding of ³H-ryanodine (Ohkusa et al., 1999). In vitro studies in normal tissues have found that activation of RyR2s with ryanodine (0.5–2 µM) results in increased spontaneous electrical activity in the PVs. As has been reported in the ventricle during HF, it has been suggested that hyperphosphorylation of RyR2s contributes to diastolic SR Ca²⁺ leak and may contribute to the initiation or maintenance of AF (Vest et al., 2005). In this study of atrial tissue from either dogs or humans with chronic AF, there was increased PKA-dependent phosphorylation of RyR2 at serine 2809, reduced FKBP12.6 bound to the RyR2 complex, and increased RyR2 channel open probability, compared to controls. In addition, treatment with JTV519 was reported to normalize AF-induced abnormalities in RyR2 function. More recently, reduced activity of SR-associated PP1 has been reported in atrial tissues from patients with chronic AF (Dobrev, 2006), which could contribute to increased phosphorylation of RyR2s. However in light of recent negative results on the role of FKBP in other disease states these findings should be viewed with caution.

7. Therapeutic strategies

7.1. Arrhythmia

Current antiarrhythmic treatments are often empiric and ineffective. One significant limitation to the effective prevention and treatment of arrhythmias is that the initiating and/or perpetuating event(s) are often unknown in a given patient. For arrhythmias of the atria or ventricles, pharmacologic antiarrhythmic therapies act by blocking fast inward sodium channels (slowing conduction) or blocking repolarizing potassium currents (prolonging refractoriness). Examples of clinically useful sodium channel blockers would include procainamide, lidocaine, mexiletine, flecainide or propafenone. Within these are drugs with pharmacology which varies from pure sodium channel block, (e.g. lidocaine), to procainamide which also prolongs refractoriness through effects on potassium channels, to propafenone which is also a non-selective beta-adrenergic blocker. Examples of clinically useful potassium current blockers include dofetilide, sotalol (also a non-selective beta-adrenergic blocker) and amiodarone (also a blocker of fast inward sodium current, slow inward calcium current, a non-selective beta-adrenergic blocker, and an antioxidant). Non-pharmacologic treatment with implantable cardioverter defibrillators has been shown to be more effective that optimal pharmacologic therapy in multiple clinical trials; although many patients receive a combination of pharmacologic and non-pharmacologic therapy for ventricular tachyarrhythmias (Carnes et al., 1998). Interestingly, the drug with the highest rate of efficacy is amiodarone which is the least pharmacologically selective.

However, for those diagnosed with specific forms of arrhythmic diseases, there may be targeted therapy. For example, catecholaminergic polymorphic ventricular tachycardia (CPVT) is typically treated initially with beta-adrenergic blockers (e.g. nadolol), with implantable cardioverter defibrillators also used for either primary or secondary prevention of cardiac arrest (Liu et al., 2007). Ideally, as arrhythmic mechanisms are better understood, additional targeted and specific therapies can be developed for the treatment and prevention of cardiac arrhythmias.

The antiarrhythmic effects of ryanodine have been investigated in multiple animal models. In canine models, ryanodine has been shown to reduce ouabain-induced (Ca²⁺ overload) arrhythmias (Lappi & Billman, 1993), but not ischemia-induced arrhythmias (Lappi & Billman, 1993). In addition, ryanodine has demonstrated efficacy in reducing early afterdepolarizations and Torsades de Pointes in a canine model of chronic AV nodal block (Verduyn et al., 1995). Similar to the findings in canines, ryanodine has been shown to reduce ouabain-induced arrhythmias in guinea pigs (Golovina et al., 1988; Zakharov et al., 1991). There is also evidence suggesting that ryanodine may reduce reperfusion-induced arrhythmias (Thandroyen et al., 1988; Hayashi et al., 1996). Thus, ryanodine, by reducing pathologic leak of Ca²⁺ from the SR, appears to have antiarrhythmic efficacy in multiple models of arrhythmogenesis, in particular when then there is increased SR Ca²⁺ load (e.g. ouabain toxicity). However, as might be expected, ryanodine is a profound negative inotrope (Thandroyen et al., 1988), and thus optimal therapy should aim to normalize, rather than simply inhibiting RyR2 function in order to maintain systolic function.

7.2. Heart failure

Heart failure is a constellation of symptoms (circulatory congestion, dyspnea, fatigue, and/or weakness) which results when cardiac output is not sufficient to meet the body's metabolic demands. One of the most common defects causing heart failure is impaired contractility of the ventricle (cardiomyopathy or ventricular systolic dysfunction). With respect to calcium cycling, hallmark characteristics of heart failure include impaired contractility, and low amplitude, broad calcium transients in isolated myocytes. Current therapies for heart failure are intended to reduce circulatory congestion, reduce cardiac hypertrophy and structural remodeling, reduce cardiac afterload and preload, and modify adrenergic signaling. Current therapeutic strategies for patients with current or previous symptoms of chronic systolic HF include a diuretic, beta-adrenergic blocker, an angiotensin converting enzyme inhibitor or angiotensin receptor blocker, and in selected patients an aldosterone antagonist (Hunt et al., 2005). Other therapeutic strategies which may be appropriate for specific patient populations include an implantable cardioverter defibrillator, cardiac resynchronization therapy, digoxin and/or combined hydralazine and nitrate therapy. The ideal therapy for HF would be one that increases contractility while preserving relaxation (lusitropy) and avoiding proarrhythmia (the development of new arrhythmias or worsening of existing arrhythmias). Given the Ca²⁺-dependence of both arrhythmia initiation and contractility this remains a difficult and challenging goal.

7.3. Normalizing ryanodine receptor function as a treatment strategy for cardiac disease

Since RyR Ca²⁺ leak contributes to various disease phenotypes ranging from uncontrolled Ca²⁺ release and SR depletion, simply blocking RyRs is not a viable strategy for treating arrythmia or HF in general. Selectivity for RyR2s (relative to the RyR1 complexes in skeletal muscle) would be critical for any clinically viable therapy targeting RyR2 modulation. Additionally, implementing a successful therapy based on RyR2 targeting could be challenging due to the complex nature of store-mediated Ca²⁺ signaling. Indeed, inhibition of RyR2s may fix the leak, but has the potential to impair systolic function. Moreover, inhibition of RyR2s can also result in inhomogeneities of Ca²⁺ release that can be arrythmogenic (Diaz et al., 2002). Still RyR inhibitors, such as an RyR2-selective derivative of tetracaine, might prove to be useful for treating certain conditions when the SR Ca²⁺ leak is excessively large such as end-stage HF. Indeed, Eisner and coworkers (Venetucci et al., 2006) showed that tetracaine increased Ca²⁺ transients by increasing the SR Ca²⁺ load in myocytes exposed to isoproterenol. A caveat is that the therapeutic action of such an agent would be highly concentration-dependent such that at high concentrations it could inhibit systolic Ca²⁺ release.

RyR2 inhibitors could be potentially used in combination with other agents to enhance SERCA-mediated SR Ca²⁺ uptake thus restoring further SR Ca²⁺ content. Although potentially useful for treating end-stage HF where the SR Ca²⁺ content is greatly diminished, both strategies could be detrimental by increasing the risk of arrhythmia. For example, tetracaine can cause homogeneities in release, while adrenergic stimulation (which occurs in response to HF) in isolation can promote arrhythmias. Given these problems with using RyR2 antagonists, targeting the underlying causes of altered RyR2 function appears to be a more promising approach and developing such therapies might be the ultimate goal in treating RyR2-linked cardiac diseases. Currently several such strategies are emerging based on specific mechanisms proposed to underpin dysregulation of RyR2s in cardiac disease, including abnormal PKA and/or CAMKII phosphorylation and/or altered redox modification.

Much attention has been devoted to a Ca²⁺ channel blocker derivative with reported beneficial effects in HF and a range of reported intracellular targets such as PKC, mitochondrial K⁺ channels, and RyR2s. It has been suggested that JTV519 improves Ca²⁺ handling in both HF and CPVT by stabilizing FKBP12.6 binding to RyR2s thereby reducing the RyR-mediated SR Ca²⁺ leak (Yano et al., 2003; Wehrens et al., 2005b; Lehnart et al., 2006). However this mode of action of JTV519 has been disputed by other investigators, further adding to the uncertainties regarding the role of FKBP12.6 in HF-related alterations in RyRs. While most studies agree that JTV519 improves cardiac function in HF, its beneficial effects in CPVT are not supported by recent investigations (Liu et al., 2006; Hunt et al., 2007; George et al., 2007). This may indicate that the beneficial effects of JTV519 on cardiac function in HF do not involve RyRs, or that the RyR defects in CPVT and HF are caused by different mechanisms.

Given the recent evidence that RyR-mediated alterations in HF and in certain types of arrhythmia are mediated by CAMK2-phosphorylation of RyR (Ai et al., 2005; Grueter et al., 2007) targeting CAMK might be another strategy to normalize RyR function in cardiac disease. Finally, preventing/reversing pathologic RyR modification by reactive oxygen/nitrogen species has been recently proposed as a strategy to treat HF and arrhythmias caused by abnormal Ca²⁺ release. The ability of antioxidants to improve cardiac performance was demonstrated in a number of studies using animal models of HF (e.g. Mochizuki et al., 2007). However, a recent clinical trial examining xanthine oxidase inhibition was not successful (Hare et al., 2006). It is unclear to what extent this lack of success is due to the particular antioxidants used. A better understanding of the specific redox pathway abnormalities and the affected Ca²⁺ signaling mechanisms may enable the development of new rationally targeted therapeutic strategies.

8. Conclusion

In recent years there has been significant progress in understanding Ca²⁺ handling in the heart and how changes in Ca²⁺ homeostasis contribute to cardiac disease. Exploring the functional consequences of naturally occurring arrhythmogenic mutations in Ca²⁺ handling proteins such as RyR2 and CASQ2 has been especially instructive. These studies have led to a better understanding of the control of SR Ca²⁺ release at the molecular level in normal and disease settings. An important question for future studies is how derangement of one aspect of Ca²⁺ handling, i.e. Ca²⁺ release through RyR2s, can result in awide spectrum of disease phenotypes ranging from arrhythmogenic Ca²⁺ alternans and dispersed SR Ca²⁺ release to depleted SR Ca²⁺ stores in HF. A comprehensive understanding of these mechanisms may provide a conceptual framework for the development of individualized, targeted therapy of cardiac diseases.

Abbreviations

AF

atrial fibrillation

AP

action potential

CAMKII

calmodulin-dependent protein kinase II

CASQ

calsequestrin

CN

calcineurin

CPVT

catecholaminergic polymorphic ventricular tachycardia

DAD

delayed afterdepolarization

DTT

dithiothreitol

EAD

early afterdepolarization

FKBP

FK binding protein

fSR

free sarcoplasmic reticulum

HF

heart failure

JN

junctin

jSR

junctional sarcoplasmic reticulum

NCX

sodium–calcium exchanger

PKA

protein kinase A

PLB

phospholamban

PV

pulmonary vein

QTc

rate-corrected QT interval of the electrocardiogram

RyR

ryanodine receptor

SERCA

sarco-endoplasmic reticulum calcium ATPase

TRD

triadin

Vm

membrane voltage